We propose to use magnetic resonance imaging (MRI) and histology to determine the time course of development of radial pathways, thalamo-cortical fibers, short- and long-range cortico-cortical fibers and cortical efferent projection fibers in the brain of the human fetus ex vivo. Amajor challenge in MR tractography is to distinguish different fiber tracts at locations where they cross one another. When two or more fiber tracts cross within a voxel, conventional diffusion tensor imaging (DTI) often fail to resolve the multiple angles with peak water diffusion. This leads to errors such as missing fiber crossings when they exist or misconnecting incorrect fiber tracts. To overcome this problem, our colleagues have developed high-angular resolution techniques. They resolve the water diffusion angles along 60-200 directions and detect fiber crossings by the presence of multiple peaks in the water diffusion spectrum. Even with these high-resolution techniques, however, there is an additional problem when one tries to apply these techniques to determine the fiber pathways in developing fetal brains. Because myelination is incomplete, and the pattern of myelination dynamically changes during development, accurate tractography on human fetal brains has been challenging. We have recently shown that this problem can be overcome with HARDI and DSI by further refining scanning conditions and data analysis technique. We are now able to perform tractography in the developing brain without using a water diffusion anisotropy threshold. This has enabled us to visualize detailed coherent structures in the developing brain even in the areas with low myelination. This R21 proposal takes advantage of these recent technical developments to address an important issue in developmental human neuroscience, namely the time course of development of fiber pathways in human fetal brains. Although this issue has been addressed by others with diffusion tractography, these studies thus far have used DTI with incomplete pathway tracking. Therefore, we still know very little about the spatio-temporal course of fiber development in human brain. We believe that our technical developments will enable us to demonstrate that important developmental issues can be addressed in ex vivo preparations. For this first systematic high-resolution study of fiber development in the whole human fetal brain, we will use postmortem samples. This approach will provide crucial normative data with high signal-to-noise ratios, since movement artifact will be completely absent. Moreover, we will be able to validate our tractography results by comparing with histology. As discussed in the proposal, we believe that this research has potential for important clinical applications in perinatal brain injury or malformation as well as understanding of subtle abnormality of folding/connectivity that affect higher cognitive functions after birth.